Recently, certain processes have been described for depositing thin films of metal chalcogenides. For example, see U.S. Pat. No. 6,875,661 to Mitzi, entitled “Solution Deposition of Chalcogenide Films”; US Patent Publication 2005-0009225 to Mitzi et al., filed Mar. 16, 2004 entitled “Hydrazine-Free Solution Deposition of Chalcogenide Films,” and US Patent Application Publication 2005-0158909 to Mitzi et al and entitled “Solution Depositon of Chalcogenide Films Containing Transition Metals,” the entire disclosures of which are incorporated herein by reference.
The thin films of the metal chalcogenides are deposited using solutions prepared by dissolving a metal chalcogenide material in a hydrazine or hydrazine-like solvent. The metal chalcogenide may be of the form MX, MX2, M2X3 or M2X where M=metal (e.g., Sn, Ge, Pb, In, Sb, Hg, Ga, Tl, K) or a combination thereof and X=chalcogen (e.g., S, Se, Te) or a combination thereof. Since the metals often have the potential for multiple oxidation states, the metal chalcogenide may often be non-stoichiometric and may therefore more generally be represented as MyXz (where 0<y, z and may be an integer or non-integer).
The above-described processes for shorthand purposes can be referred to as hydrazine-precursor techniques or processes. The hydrazine-precursor technique has the advantage of being a high-throughput process, which does not require high temperatures or high vacuum conditions for the thin-film deposition process. The hydrazine precursor process thereby has the potential for being low-cost and suitable for deposition on a wide range of substrates, including those that are flexible. As metal chalcogenides can exhibit a wide range of electronic character, it may be used to prepare high-quality semiconducting, insulating or metallic films. The process has been used to deposit, for example, both n- and p-type semiconducting films for use as channel layers in thin-film transistors (TFTs), exhibiting field-effect mobilities>10 cm2N-s—approximately an order of magnitude better than previous results for spin-coatable semiconductors [“High Mobility Ultrathin Semiconducting Films Prepared by Spin Coating, Nature, vol. 428, 299 (2004)].
Besides TFTs, other electronic devices that rely on metal chalcogenide films can also be prepared using the described technique. Solar cells, for example, may consist of thin n-type chalcogenide semiconductor layers (˜0.25 μm) deposited on a p-type substrate, with electrical contacts attached to each layer to collect the photocurrent. Light-emitting diodes (LEDs) are typically comprised of a p-n bilayer, which under proper forward bias conditions emits light.
Rewriteable phase-change memory generally employs a film of a chalcogenide-based phase-change material, which must be switchable between two physical states (e.g., amorphous-crystalline, crystalline phase 1-crystalline phase II). The state of the phase change material must also be detectable using some physical measurement (e.g., optical absorption, optical reflectivity, electrical resistivity, index of refraction). As an example, commercially-available rewritable optical memory generally relies on a film of a metal chalcogenide material such as Ge2Sb2Te5 or KSb5S8 [“KSb5S8: A Wide Bandgap Phase-Change Material for Ultra High Density Rewritable Information Storage,” Adv. Mater., vol. 15, 1428, 2003]. Initially the film is amorphous, but may be converted to a crystalline form using a laser beam of sufficient intensity to heat the material above the crystallization temperature. Subsequent exposure to a more intense and short laser pulse melts the crystallized chalcogenide phase-change material, resulting in a conversion to an amorphous state upon quenching. A recorded bit is an amorphized mark on a crystalline background. The reversibility of the crystallization-amorphization process allows for the fabrication of rewritable memory [A. V. Kolobov, “Understanding the phase change mechansim of rewritable optical media, Nature Mater., vol. 3, 703, 2004].
Generally the chalcogenide materials in the above-described applications are deposited using vacuum-based techniques such as sputtering or thermal evaporation. A solution-based process is desirable because of the reduced complexity of the process (reducing cost and improving throughput) and the ability to deposit on a wider range of substrate types (including those that have very large area or are flexible) and surface morphologies.
One disadvantage of the above-described hydrazine-precursor process is that it requires the isolation of the metal chalcogenide before the deposition-process can be initiated. Metal chalcogenides are often formed using high-temperature (energy-intensive) and/or multi-step (time-consuming) reactions. Given, the prevalence of multiple possible compositions for a given M and X, the formation of single phase metal chalcogenide starting materials can also be problematic. As an example, tin sulfide can exist as SnS or as SnS2, or perhaps more appropriately as SnS2-x to accomodate the potential for non-stoichiometry [“Preparation and Characterization of SnS2,” J. Solid State Chem., vol. 76, 186]. The reaction of a 1:2 (molar) ratio of Sn and S at high temperature often yields SnS2, in addition to impurities of SnS and S. The use of these SnS2 materials for thin-film deposition may therefore lead to non-reproducibility since the exact composition of the starting metal chalcogenide may vary from run to run.